This revised proposal requests support for a three-year study of stimulus control and transitive processes in mentally retarded individuals' discrimination learning. The project seeks to clarify learning processes involved in the formation of stimulus classes defined by ordinal stimulus relations (i.e., sequence classes). Comparisons of several training methods will identify behavioral histories that lead to acquisition of ordinal relations that are also transitive. Our primary concern is to determine whether new ordinal relations will emerge without direct training via transitive relations among sequence-class members. We expect to show that, given appropriate learning histories, transitive performances are demonstrable in persons functioning at much lower developmental levels than generally thought. We also expect to demonstrate that the behavioral processes to be investigated are formally similar and highly relevant to those involved in acquisition of rudimentary academic behaviors, such as arithmetic and money skills. Our research is likely, therefore, to contribute to a growing technology for teaching mentally retarded people and others with special instructional needs.